Download Transaction Logs

1
Chapter 1
INTRODUCTION
Choosing a database management system (DBMS) for any application is an
important decision for developers because their applications need to run efficiently. This
decision should be based on the information that compares different database
management systems based on various performance factors. Different applications have
unique characteristics in their functionalities. Application developers need to choose
appropriate technologies for the development of their product. This project will help them
make a decision for the DBMS selection. The project compares two different database
management systems based on some performance factors under various constraints.
The main purpose for this project is to analyze and compare the performance of
two widely used database management systems. One of them is SQL and the other is
Oracle. Both technologies have been implemented under various test conditions and their
behaviors analyzed.
The measurement factors are Memory Requirements and Execution time for
performing various types of queries. These measurements have been taken on same
database schema for the both technologies. The same sets of queries have been analyzed
by gradually increasing the number of records in the database. Performance impact of
size increment have also been measured and compared. Performance data for both
technologies are displayed on line charts for ease of comparison.
2
The targeted end users of this paper are application developers who work in
development environments where a database plays an important role as a backend and for
whom the performance is an important aspect. They can review the data of project
results as well as the conclusion to determine which database management system to use
for their applications.
3
Chapter 2
BACKGROUND
During the initial stage of this project, a literature review was done to get an
understanding of SQL and Oracle. Various papers were reviewed [1], [6], [7], [9], [10],
[11] to identify the performance factors to be analyzed. In the beginning, measurement
factors such as Failure Behavior, Failure Recovery, Query Execution Time, Portability,
Scalability, Memory Usage, Transaction Support, Multiuser Support and some others
were proposed. After considering them there was a need for elimination of some of the
proposed factors to make the scope of the project feasible. Review of the literature helped
to finalize two performance factors: Query Execution Time and Memory Requirements
for study in this project. These two factors have been considered by focusing on the needs
of industry of efficient and cost effective applications.
This literature review helped in the understanding of the methods for measuring
and comparing the performance of both systems. It also helped to get the understanding
of enforcing constraints, such as, identical hardware configurations and system
environments. Based on the literature review, decision on the design of the database
schema, the numbers of records in the database, and the query types to be analyzed were
made.
The deciding factors in this project have been efficient time and memory in
database query applications.
4
Chapter 3
SQL DATABASE MANAGEMENT SYSTEM
3.1 SQL Background
SQL often referred to as Structured Query Language, is a database computer
language designed for managing data in relational database management systems. The
scope of SQL includes query, data insert, delete, update, schema creation and
modification, and data access control.
The SQL language is sub-divided into several language elements, including:
i.
Clauses: They are components of statements and queries.
ii.
Expressions: They can produce either scalar values or tables consisting of
columns and rows of data.
iii.
Predicates: They are used to limit the effects of statements and queries, or to
change program flow.
iv.
Queries: They retrieve the data based on specific criteria. This is the most
important element of SQL.
v.
Statements: They may have a persistent effect on schemas and data, or which may
control transactions or program flow.
5
3.2 Queries
For this project we will focus on Queries.
Queries
The most common operation in SQL is the query which is performed with the
declarative SELECT statement. SELECT retrieves data from one or more tables.
Standard SELECT statements have no persistent effects on the database. Some non
standard implementations of SELECT can have persistent effects, such as the SELECT
INTO syntax that exists in some databases. Query allows the user to describe desired
data, leaving the database management system responsible for planning, optimizing and
performing the physical operations necessary to produce that result.
A query includes a list of columns to be included in the final result following the
SELECT keyword. An asterisk can also be used to specify that the query should return all
columns of the table. The optional keywords and clauses for SELECT statement in SQL
include:
i.
The FROM clause which indicates the table from which data is to be retrieved.
ii.
The WHERE clause includes comparison predicate, which restricts the rows
returned by the query.
iii.
The GROUP BY clause is used to project rows having common values into a
smaller set of rows.
6
iv.
The HAVING clause includes a predicate used to filter rows resulting from the
GROUP BY clause.
v.
The ORDER BY clause identifies which columns are used to sort the resulting
data, and in which direction they should be stored. Without ORDER BY clause,
the order of rows returned by an SQL is undefined.
3.3 Advantages of SQL
Stored Procedures
The main advantage of using SQL Server is the use of stored procedures. Stored
procedures are lines of code that are called by the application. They are placed on the
server, and they are pre-compiled for quicker response times. Stored procedures require
the knowledge of SQL Server syntax, which is called T-SQL. The use of stored
procedures also centralizes code, so troubleshooting bad database requests can be
observed by a database administrator.
Scalability
The term scalability is used to describe the ability to grow when the business
becomes bigger. Microsoft SQL server is quick for large and small businesses, so as the
business grows, the SQL Server can handle the new volume of database requests. SQL
server can handle millions of records and transactions.
7
Security
Security is a major issue for any site. SQL Server allows the administrator to
grant access or deny access for users. The SQL Server has a specific section of the
application where users are added to the permissions. SQL Server allows administrators
to specify which tables and stored procedures users are able to access and query. This
limits what records and user information can be queried.
Transaction Logs
Transaction logs are objects on the SQL Server that record the retrieval, update
and deletion of records. There are two reasons to keep transaction logs. The first is for
rollback procedures. This process is used for accidental updates or deletions. The
administrator can return records back to the original data by using transaction logs.
Secondly, the transaction logs can be used for security purposes. If the administrator
suspects a breach of security, he can watch the transaction logs for any type of data
retrieval and identify the severity of the breach.
Automatic Backup
SQL Server has an automatic backup option. The SQL Server automatically saves
a copy of the database and the transaction logs on another hard drive or media like a CD-
8
ROM or a DVD. SQL Server also has procedures that allow the administrator to quickly
restore a database when data is lost or corrupted, or the server has a hard drive crash.
3.4 Problems of SQL
Some of the problems that may occur [15] with SQL are described below.
i.
The language makes it too easy to do a Cartesian join, which results in "runaway" result sets when WHERE clauses are mistyped.
ii.
The grammar of SQL is perhaps unnecessarily complex, borrowing a COBOL
(Common Business Oriented Language) like keyword approach, when a functioninfluenced syntax could result in more re-use of fewer grammar and syntax rules.
iii.
Implementations are inconsistent with the standard and, usually, incompatible
between vendors. In particular date and time syntax, string concatenation, NULLs
and comparison case sensitivity vary from vendor to vendor.
iv.
Inability to implement Recursive Processing: One of the major drawbacks of SQL
is its incapability to execute recursive processing. Recursive processing is a type
of computer function in which one of the steps of a program reruns the entire
program. SQL lacks looping constructs that are common in other high-level
programming languages, such as for and while loops. It cannot repeat actions and
there is no way to define repetitive looping constructs in SQL.
9
Chapter 4
ORACLE DATABASE MANAGEMENT SYSTEM
4.1 ORACLE Background
The Oracle Database [16], [17], [18], [19] is an object-relational database
management system (ORDBMS). The Oracle RDBMS stores data logically in the form
of table-spaces and segments, such as Data Segments, Index Segments, etc. Segments, in
turn, comprise one or more extents. Extents are comprised of groups of contiguous data
blocks. Data blocks create the basic units of data storage.
Oracle database management tracks its computer storage with the help of
information stored in the SYSTEM table-space. The SYSTEM table-space contains the
data dictionary and often indexes and clusters. A data dictionary consists of a special
collection of tables that contains information about all user-objects in the database.
Each Oracle instance uses a system Global Area or SGA – a shared memory area
– to store its data and control information. Oracle instance allocates itself an SGA when it
starts and de-allocated it at shut-down time. The information in the SGA consists of the
following elements, each of which has a fixed size, established at instance startup:
10
i.
The redo log buffer: It stores redo entries—a log of changes made to the database.
The instance writes redo log buffers to the redo log as quickly and efficiently as
possible. The redo log aids in instance recovery in the event of a system failure.
ii.
The shared pool: This area of the SGA stores shared-memory structures such as
shared SQL areas in the library cache and internal information in the data
dictionary. An insufficient amount of memory allocated to the shared pool can
cause performance degradation.
4.2 ORACLE Advantages
Oracle is used for almost all large applications and one of the main applications in
which Oracle takes its major presence is banking. Oracle offers a powerful combination
of technology and comprehensive, pre-integrated business applications, including key
functionality built specifically for banks. Oracle frequently upgrades and releases new
products into market.
Oracle is a database that responds very well with excellent performance in
demanding environments. Oracle is a major database which along with its added features
passes the ACID (A-Atomicity, C-Consistency, I-Isolation, D-Durability) test:
11
i.
Atomicity: The the result of a transaction's execution are either all committed or
all rolled back.
ii.
Consistency: The database is transformed from one valid state to another valid
state. Illegal transactions aren't allowed and, if an integrity constraint can't be
satisfied then the transaction is rolled back.
iii.
Isolation: The results of a transaction are invisible to other transactions until the
transaction is complete, thus, increasing the security on data.
iv.
Durability: Once committed, the results of a transaction are permanent and
survive future system and media failures and thus ensuring maintenance and
protection of data.
4.3 ORACLE Disadvantages
Some of the limitations of Oracle are [17], [18]:
i.
Connection Takes More Memory: Each database connection takes 2-4 MB of
memory on the host machine where as for SQL the amount of memory is less than
1/10th of that.
12
ii.
Incompatibility and Complexity: One of the major disadvantages of Oracle is
inconsistency and data incompatibility in the areas of time and date syntax, string
concatenation and case sensitivity. The language is complex, with a keyword
approach similar in structure to COBOL (Common Business Oriented Language),
with fewer syntax and grammar rules.
iii.
Oracle doesn’t support Update within Join: With Oracle you can’t run a query
like this:
“ update history set a.teller_id = h.teller_id
from account a, history h
where a.account_id = h.account_id ”
In Oracle, you need to use convoluted sub-select format, to achieve the same
result.
iv.
There is no database dump utility: There is no “block_dump” style backup
program in Oracle. Space reclamation is also poor. In Oracle, tables grow even
after deletion of row. A query might scan empty blocks of deleted rows.
13
Chapter 5
PROJECT PROBLEM AND SOLUTION APPROACH
5.1 Purpose and Goal
For any application which has database as a backend, data retrieval time and
memory usage are crucial performance factors. So choosing a right DBMS technology is
an important decision for any application developer. This project is designed to help users
make the right decision. Behavior of both SQL and Oracle will be analyzed under various
test conditions. Measuring performance of DBMS for different types of query sets can
help determine which one is better for what types of queries. Also, there will be
measurements for memory usages.
To achieve this goal, there will be two systems with SQL and Oracle installed on
each computer. The database schema will be designed for both DBMS. Selected types of
queries will be executed and their performance in terms of time and memory usage will
be analyzed. Analysis will be repeated every time when some fixed numbers of records
have been inserted into both databases. There are also some constraints that will be
applied for accurate measurements.
14
5.2 Project Requirements and Constraints
The requirement for this project will be two working computers having same
hardware configuration and same operating system environment. In comparing two
different systems, the main constraint of this project is to maintain the same resources
and environment for both of them. Two database schemas, one for Oracle and other one
for SQL were created.
5.3 Database Schema
The implementation phase started with designing database schema for analysis.
As we are comparing two different technologies it is required that both of them have the
same database schema. By that we can provide a consistent base for both DBMSs. Here I
have designed the database of bank account. The schema for the same is given in the
figure below.
15
branches (BranchId, BranchBalance)
tellers (TellerId, BranchId, TellerBalance)
accounts (AccountId, AccountBalance, BranchId)
history(AccountId,
Amount, Date)
Figure 1:TellerId,
DatabaseBranchId,
Schema Design
Figure 1 Database Schema Design
The above diagram shows the Database Schema design for both DBMS. Names
on the left of parentheses are table names. The names within parentheses are columns of
those tables. Underlines specify primary key of the table. Each arrow is for the foreign
key which is being referenced from parent table. We can see there are four tables in this
database.
i.
BRANCHES: This table contains branch specific details. It has two columns:
“BranchId” which is a primary key of the table and contains unique identification
16
number for specific branch. Another is “BranchBalance” which specifies total
amount particular branch is holding.
ii.
TELLERS: This table is having details for a specific teller. It has three columns:
“TellerId” which is a primary key for the table and contains unique identification
number for a teller. Others are “BranchId” and “TellerBalance”. BranchId is the
foreign key referenced from table Branches. It specifies the branch that a teller
works for. “TellerBalance” is the total amount the teller has in his/her account.
iii.
ACCOUNTS: This table contains account related information. It has three
columns: “AccountId”, “AccountBalance” and “BranchId”. “AccountId” is a
primary key for the table. It contains unique identification number of any account.
“AccountBalance” is the current balance of the account for that “AccountId”.
“BranchId” is the foreign key referencing the table Branches. It describes the
branch under which the account exists.
iv.
HISTORY: This table describes transaction that has occurred in the past stating
the amount and date of its occurrence. It is having “AccountID”, “TellerId” and
“BranchId” as its primary key. Together they all describe transaction that has
occurred in the account of which branch and under which teller. Other two are
“Amount” and “Date”. They specify amount of the transaction with the date on
which they occurred. It should be noted that together “BranchId”, “AccountId”
and “TellerId” are primary key. But individually each of them are foreign key
referencing tables Branches, Accounts and Tellers accordingly.
17
5.4 ER-Diagram
Next step in implementation was to draw ER-Diagram based upon developed
database schema. Figure drawn below shows the ER-Diagram for the same.
branches
BranchId
BranchBalance
e
1
1
accounts
_of
teller_of
history
N
N
Amount
Date
tellers
TellerID
accounts
TellerBalance
AccountId
Figure 2: Database ER Diagram
AccountBalance
18
In above figure, each rectangle stands for entities. Texts in them are the table
names that particular square is representing. Ovals represent attributes of entities.
Diamonds stands for relation between entities. Attribute names which are underlined
represent primary key. Lines represent relation between connected objects. For example
attributes “BranchId” and “BranchBalance” are linked to entity “Branches”. Relation
“History” is ternary relation as it is connected to three different entities which are
“Branches”, “Accounts” and “Tellers”. Whereas other two relations “Accounts_of” and
“Teller_of” are binary relations as they are relating two entities.
The numbering on the connecting lines shows cardinality of relation. For relation
between “Branches” and “Accounts”, the cardinality is 1:N. That means one branch can
have many accounts but a single account can be under one branch only. Same concept
applies to relation between “Branches” and “Tellers” also.
5.5 DBMS Installation and Database Creation
After designing database, the next task was to install DBMSs in two different
computers. Both computers have the same hardware configuration. Oracle and SQL have
been installed on them. For Oracle, the software is “Oracle Database 10g Express
Edition” and for SQL, “SQL Server 2008 R2” has been installed. It was obtained for free
from their vendor’s website.
19
After installation, the next task was to start creating a database on each system.
When that was done, the memory space required for each database on both DBMSs has
been measured. After successfully creating databases, data insertion was started and
different types of queries have been implemented on them.
5.6 Finalizing Queries for Analysis
Before Implementation, one task that needs to be done is to choose types of
queries for analysis. The research [11] that I did in the beginning of the project helped me
to finalize some types of queries important for performance analysis. Among different
types, five types of queries were selected. They are:
i.
Insert Query: There are four Insert queries for both databases, one query for each
of the four tables. They are used to insert records into the database.
ii.
Select Query: Same as Insert, there are four Select queries for both databases, one
for each of the four tables. It will display all the records of the table from the
database.
iii.
Join Query: There is one Join query that links all four tables of database and
displays the result.
iv.
Update Query: There are four Update queries, one for each of the four tables.
Update Query can change any single row anywhere in database that satisfies
given condition.
20
v.
Delete Query: Again, four queries. One for each of the four tables. Delete query
removes the entire row that satisfies the given condition.
5.7 Query Execution and Analysis
After the database was implemented and the types of queries finalized
performance analysis was conducted. In this phase, one row is inserted at a time into the
table. When inserted row counts reaches fifty, performance analysis is done by executing
all five types of queries mentioned in the previous section. After completing analysis,
again records insertion continues. When row counts reaches hundred i.e. another fifty
rows have been inserted, again query performance evaluation begins. So in this project,
records have been inserted in blocks of fifty rows for each table. At every row count of
fifty for each table i.e. two hundred per database, query performance analysis is
performed. This analysis continues until the insertion of seven hundred rows per table.
That means 2800 rows per database, which means total of 9100 records. There are two
similar databases so the total analysis is for 5600 rows, i.e., 18,200 records.
When a query is executed, the results are checked for correctness. For each
query, execution time or memory usages are measured depending upon query type. All
the results of the measurements are described in next chapter.
21
Chapter 6
PROJECT RESULTS
This chapter discusses the results of the performance measurements that have
been obtained during performance analysis.
6.1 Insert Query Results
As discussed in the previous chapter there are four different insert queries per
database – one for each database table. Sample queries for inserting one row in each table
are:
i.
For Table “BRANCHES” : insert into branches(BranchId,BranchBalance)
values('b0001',1000001)
ii.
For Table “TELLERS”: insert into tellers(TellerId,BranchId,TellerBalance)
values(‘t0100’,’b0100’,’10000100)
iii.
For Table “ACCOUNTS”: insert into
accounts(AccountId,AccountBalance,BranchId) values('a0700',100600,'b0700')
iv.
For Table “HISTORY”: insert into
history(AccountId,TellerId,BranchId,Amount,Date1) values
('a0700','t0700','b0700',1600,'02/Mar/11')
22
Below is a line chart describing Insert Query behavior of the both technologies as
number of rows increases. Data of History table have been selected for the presentation
purpose. Actual data of results are displayed in Appendix. There are considerable
differences in execution time between SQL and Oracle. SQL takes more time to execute
a query than Oracle. We can see that in the chart that for SQL, there are peaks in
execution time due to additional load. At the same time, Oracle is able to absorb the
additional load successfully without impacting the performance in most of the cases.
0.25
0.2
0.15
Oracle
0.1
SQL
0.05
0
Figure 3: Insert Query Behavior Chart
23
6.2 Select Query Results
As discussed before Select query displays the results obtain from the table. Like
Insert query, there are four select queries per database for each table. Sample select
queries for each table are given below.
i.
For Table “BRANCHES” : select * from branches
ii.
For Table “TELLERS”: select * from tellers
iii.
For Table “ACCOUNTS”: select * from accounts
iv.
For Table “HISTORY”: select * from history
Below is a line chart showing Select query behavior of the both technologies as
number of rows increases. Actual measurements are given in Appendix. Data of History
table have been selected for the presentation purposes. From this chart, it can be seen that
for Select query, Oracle takes considerably less time than SQL. Peaks in the lines show
the performance impact due to additional loads on both systems.
24
0.2
0.18
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
Oracle
SQL
Figure 4: Select Query Behavior Chart
6.3 Join Query Results
Join query displays selected records from all four tables of a database which
satisfies the condition in the given query. There is just a single join query per database in
this project. Sample query for both DBMS is:
select * from accounts a, branches b, tellers t, history h where
h.AccountId=a.AccountId and h.TellerId=t.TellerId and h.BranchId=b.BranchId;
Below is a line chart describing Join query behavior of both technologies as
number of rows increases. Data is provided in the Appendix. Data from History table is
25
provided for presentation purposes. As before, Oracle requires less time than SQL for
join queried.
0.45
0.4
0.35
0.3
0.25
0.2
Oracle
0.15
SQL
0.1
0.05
0
Figure 5: Join Query Behavior Chart
6.4 Update Query Results
Update query updates any single row of the table if the row satisfies the given
condition of the query. There are four update queries for a database, separate for each
table. Below are the sample queries for each table.
i.
For Table “BRANCHES” : update bank.dbo.branches SET BranchBalance=99
where BranchId='b0650'
26
ii.
For Table “TELLERS” : update bank.dbo.tellers SET TellerBalance=99 where
TellerId='t0650'
iii.
For Table “ACCOUNTS”: update bank.dbo.accounts set AccountBalance=99
where AccountId='a0650'
iv.
For Table “HISTORY”: update bank.dbo.history set Amount=99 where
AccountId='a0650' and TellerId='t0650' and BranchId='b0650'
Below is a line chart describing Update Query behavior of the both technologies
as number of rows increases. Selected data of HISTORY table is presented below. It can
be seen from the chart that the performance of the both systems are almost same in
normal situations. When there is an additional load on the system, performance of SQL is
impacted more negatively than for Oracle in some cases. But still the overall performance
is same.
27
0.16
0.14
0.12
0.1
0.08
Oracle
0.06
SQL
0.04
0.02
0
Figure 6: Update Query Behavior Chart
6.5 Delete Query Results
Delete query removes single row which satisfies enforced condition. There are
four delete queries for four databases. The sample delete queries for each table are given
below.
i.
For table “BRANCHES”: delete from branches where BranchId='b0700'
ii.
For table “TELLERS”: delete from tellers where TellerId='t0700'
iii.
For table “ACCOUNTS”: delete from accounts where AccountId='a0700'
iv.
For table “HISTORY”: delete from history where AccountId='a0700' and
TellerId='t0700' and BranchId='b0700'
28
Below is a line chart describing Delete Query behavior of the both technologies as
number of rows increases. Selected data of HISTORY table are presented. It can be seen
that for Delete query, there is not much difference in performance for most of the cases.
Peaks in the chart are performance impact of additional load. It can be seen at some point
that Oracle is affected negatively by an increased load than SQL.
0.35
0.3
0.25
0.2
Oracle
0.15
SQL
0.1
0.05
0
Figure 7: Delete Query Behavior Chart
29
6.6 Memory Usage
This section discusses the memory requirements for database tables as the number
of rows increases. For SQL there is a built-in tool to measure the memory usage. For
Oracle, in addition to the built-in tool, queries were used to get the memory usages. Data
for actual memory usage for both SQL and Oracle are given in the tables in Appendix.
Below is a chart describing Memory Usage of the both technologies as number of rows
increases. Memory usage for Oracle is constant while for SQL it increases in proportion
to number of rows in table. Still memory usage for Oracle is very high than for SQL.
0.07
0.06
0.05
0.04
Oracle
0.03
SQL
0.02
0.01
0
Figure 8: Memory Usage Behavior Chart
30
Chapter 7
COMPARISON OF SQL AND ORACLE
This chapter discusses the results obtained and analyzes which DBMS is
appropriate for what types of queries and for memory requirements.
7.1 Insert Query Analysis
After reviewing tables in section 5.1 one can say that Oracle provides faster
response time than SQL in case of Insert Query. For all four tables of database average
query response time for Oracle is far better than SQL for the same Query. Table below
describes average response time for all four tables for both DBMS. There is also a chart
available on next page for the same data.
Oracle
SQL
BRANCHES
7.7 ms
34.8 ms
TELLERS
2.1 ms
19.6 ms
ACCOUNTS
2.1 ms
25.6 ms
HISTORY
3.2 ms
29.0 ms
Table 1: Average Execution Time for Insert Query
31
Figure below specifies the column chart for the same data in the table above.
40
35
30
25
Oracle
20
SQL
15
10
5
0
BRANCHES
TELLERS
ACCOUNTS
HISTORY
Figure 9: Average Execution Time for Insert Query
32
7.2 Select Query Analysis
Reviewing tables in section5.2 gives an idea that in select also Oracle gives better
performance than SQL. Again for all four tables Oracle has given better performance
than SQL. Table below compares the average response time for both technologies for
same select query. There is also a chart available on the next page for the same data.
Oracle
SQL
BRANCHES
22.4 ms
120.1 ms
TELLERS
27.9 ms
93.0 ms
ACCOUNTS
20.3 ms
101.1 ms
HISTORY
39.1 ms
112.3 ms
Table 2: Average Execution Time for Select Query
33
Figure below describes the column chart for the same data available in the above
table.
140
120
100
80
Oracle
SQL
60
40
20
0
BRANCHES
TELLERS
ACCOUNTS
HISTORY
Figure 10: Average Execution Time for Select Query
34
7.3 Join Query Analysis
Again after going through section 6.3, it is proved that Oracle is better responsive
than SQL in terms of time. The join query that has been used for analysis relates all four
tables and returns records satisfying conditions. Below is a comparison of average
response time for both Oracle and SQL. There also a chart available for the same data.
Response Time
Oracle
SQL
72.4 ms
209.3 ms
Table 3: Average Execution Time for Join Query
Figure below describes the chart for the data available in the above table.
Response Time
250
200
150
Response Time
100
50
0
Oracle
SQL
Figure 11: Average Execution Time for Join Query
35
7.4 Update Query Analysis
Unlike previous three types of queries for Update query, performance results for
both DBMS were two sided. For one table SQL gave better performance where as for
others Oracle won the race. Difference in performance is not as huge as we have seen
before for all tables. Below is the table describing fact. There is also a chart available for
the same data.
Oracle
SQL
BRANCHES
16.6 ms
23.4 ms
TELLERS
47.1 ms
18.3 ms
ACCOUNTS
11.2 ms
27.0 ms
HISTORY
18.1 ms
35.6 ms
Table 4: Average Execution Time for Update Query
36
Figure below describes the column chart for the data available in the above table.
50
45
40
35
30
Oracle
25
SQL
20
15
10
5
0
BRANCHES
TELLERS
ACCOUNTS
HISTORY
Figure 12: Average Execution Time for Update Query
37
7.5 Delete Query Analysis
For delete query also the results were mixed. For some tables Oracle was better
while for some other SQL was impressive. And for some they both reacted almost same.
The reason for such behavior is that in situation of applying an additional load, SQL
performs better and Oracle is impacted considerable. In normal situation i.e. without any
additional load, Oracle performs well. So it can be stated that for delete query, both
technologies are almost same. Below is the table, comparing response time for both SQL
and Oracle. There is also a chart available on next page for the same data.
Oracle
SQL
BRANCHES
1.4 ms
32.2 ms
TELLERS
167.4 ms
28.5 ms
ACCOUNTS
34.4 ms
66.7 ms
HISTORY
37.4 ms
36.1 ms
Table 5: Average Execution Time for Delete Query
38
Figure below describes the column chart for the data available in the above table.
180
160
140
120
100
Oracle
80
SQL
60
40
20
0
BRANCHES
TELLERS
ACCOUNTS
HISTORY
Figure 13: Average Execution Time for Delete Query
39
7.6 Memory Usage Analysis
For memory usage, simply the winner is SQL. One can easily tell that after going
through all the tables of Appendix. For Oracle, every table requires space of 0.06 MB.
While at the same time in SQL it begins with 0.008 MB and goes till .031 MB. That’s
still far less than what oracle is using. So it can be stated that SQL is more memory
efficient. Below are a table and a chart which are comparing the average memory usage
of both DBMS.
Oracle
SQL
BRANCHES
0.06 MB
0.011 MB
TELLERS
0.06 MB
0.013 MB
ACCOUNTS
0.06 MB
0.013 MB
HISTORY
0.06 MB
0.017 MB
Table 6: Average Memory Usage for both DBMS
40
Figure below describes the column chart for the data available in the above table.
0.07
0.06
0.05
0.04
Oracle
SQL
0.03
0.02
0.01
0
BRANCHES
TELLERS
ACCOUNTS
HISTORY
Figure 14: Average Memory Usage for both DBMS
41
Chapter 8
CONCLUSION
In analyzing the results given in the tables, it’s obvious that SQL is far better than
Oracle in terms of memory requirements. On average Oracle tables require around five
times more memory that what SQL tables need. So for the applications in which memory
usage is a concern, SQL is a better choice over Oracle based on performance analysis for
this project.
Both SQL and Oracle keep aside a chunk of memory for records in database
tables. The reason that Oracle requires more memory than SQL is that for Oracle, the size
of that chunk is 0.06 MB. So every time when there is a need for additional memory for
database tables, Oracle occupies a new memory chunk of the same size. SQL also works
the same way but the memory chunk that it asks for is much smaller, 0.008 MB which is
almost 10 times less than what Oracle requires. So every time one memory slot fills up
and new one is needed, Oracle will occupy 10 times more memory than SQL. The
resulting size of an Oracle database will be much larger than that for a SQL database.
In the case of query execution time, though, results were balanced for some query
type, Oracle performs well in most of the cases with huge timing differences. For Insert
Select and Join queries there is no comparison between them as Oracle’s execution time
is far better than SQL. For Delete query type SQL is somewhat better than Oracle for
some tables. But still overall performance is nearly same. Same for Update query. The
42
results were balanced. But if one tries to see the big picture, Oracle is better in terms of
query execution time. So for time critical applications, Oracle can be a wise selection
over SQL.
Another performance impact that one can observe in the line charts of chapter 6 is
that whenever there is an additional load is added to the system, query execution time for
SQL increases multiple times. Under similar conditions, Oracle also slows down but the
impact is not as major as for SQL in most of the cases. So one can say that Oracle is
pretty good absorbing unexpected loads than SQL. The average execution time for SQL
is also considerably higher than for Oracle due to the slowdown of SQL in such cases.
That is why one can see some huge differences between column charts of SQL and
Oracle.
For this project, two DBMSs have been analyzed, but there are other products are
available in the market. One can select them for same types of comparisons. Even one
can carry out analysis based on other parameters like security, failure behavior,
transaction property, and portability.
43
APPENDIX
Measurement Results of Performance Analysis
This section contains data obtained for each table for both databases. Line charts
in Chapter 6 and column charts in Chapter 7 are based on these results only.
There are two columns in each table. First column describes the numbers of rows
in the table for which the analysis has been made. Second column stands for actual
measurements that have been made during the project. There are a total of sixteen
measurements for each table. Measurements were taken after every block of fifty rows
had been inserted into the database.
44
Tables shown below specify the results of both DBMS for Insert query.
Insert Query Results for table “BRANCHES” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.020 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.020 Seconds
0.020 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.001 Seconds
Insert Query Results for table “BRANCHES” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.024 Seconds
0.014 Seconds
0.016 Seconds
0.016 Seconds
0.030 Seconds
0.014 Seconds
0.064 Seconds
0.045 Seconds
0.058 Seconds
0.013 Seconds
0.110 Seconds
0.018 Seconds
0.098 Seconds
0.013 Seconds
0.014 Seconds
0.045 Seconds
45
Insert Query Results for table “TELLERS” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
Insert Query Results for table “TELLERS” in SQL
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.013 Seconds
0.021 Seconds
0.015 Seconds
0.019 Seconds
0.013 Seconds
0.015 Seconds
0.015 Seconds
0.027 Seconds
0.014 Seconds
0.013 Seconds
0.031 Seconds
0.014 Seconds
0.013 Seconds
0.035 Seconds
0.013 Seconds
0.025 Seconds
46
Insert Query Results for table “ACCOUNTS” in Oracle
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
Insert Query Results for table “ACCOUNTS” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.014 Seconds
0.023 Seconds
0.018 Seconds
0.016 Seconds
0.021 Seconds
0.016 Seconds
0.020 Seconds
0.057 Seconds
0.014 Seconds
0.013 Seconds
0.043 Seconds
0.037 Seconds
0.039 Seconds
0.042 Seconds
0.018 Seconds
0.025 Seconds
47
Insert Query Results for table “HISTORY” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.020 Seconds
Insert Query Results for table “HISTORY” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.015 Seconds
0.192 Seconds
0.015 Seconds
0.018 Seconds
0.013 Seconds
0.018 Seconds
0.019 Seconds
0.030 Seconds
0.014 Seconds
0.020 Seconds
0.034 Seconds
0.014 Seconds
0.028 Seconds
0.028 Seconds
0.026 Seconds
0.030 Seconds
48
Tables given below specify the results of Select query for both DBMS.
Select Query Results for table “BRANCHES” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.001 Seconds
0.010 Seconds
0.020 Seconds
0.010 Seconds
0.020 Seconds
0.006 Seconds
0.030 Seconds
0.050 Seconds
0.080 Seconds
0.030 Seconds
0.080 Seconds
Select Query Results for table “BRANCHES” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.072 Seconds
0.094 Seconds
0.105 Seconds
0.137 Seconds
0.145 Seconds
0.139 Seconds
0.084 Seconds
0.150 Seconds
0.262 Seconds
0.165 Seconds
0.099 Seconds
0.114 Seconds
0.075 Seconds
0.080 Seconds
0.111 Seconds
0.080 Seconds
49
Select Query Results for table “TELLERS” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.020 Seconds
0.020 Seconds
0.010 Seconds
0.060 Seconds
0.050 Seconds
0.030 Seconds
0.030 Seconds
0.030 Seconds
0.090 Seconds
0.050 Seconds
0.040 Seconds
Select Query Results for table “TELLERS” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.063 Seconds
0.072 Seconds
0.072 Seconds
0.086 Seconds
0.146 Seconds
0.080 Seconds
0.111 Seconds
0.092 Seconds
0.086 Seconds
0.132 Seconds
0.121 Seconds
0.076 Seconds
0.074 Seconds
0.090 Seconds
0.103 Seconds
0.100 Seconds
50
Select Query Results for table “ACCOUNTS” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.020 Seconds
0.020 Seconds
0.020 Seconds
0.030 Seconds
0.030 Seconds
0.020 Seconds
0.040 Seconds
0.050 Seconds
0.030 Seconds
0.030 Seconds
0.030 Seconds
Select Query Results for table “ACCOUNTS” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.081 Seconds
0.073 Seconds
0.087 Seconds
0.132 Seconds
0.128 Seconds
0.080 Seconds
0.111 Seconds
0.094 Seconds
0.090 Seconds
0.083 Seconds
0.100 Seconds
0.077 Seconds
0.121 Seconds
0.075 Seconds
0.137 Seconds
0.124 Seconds
51
Select Query Results for table “HISTORY” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.010 Seconds
0.030 Seconds
0.030 Seconds
0.080 Seconds
0.030 Seconds
0.080 Seconds
0.030 Seconds
0.160 Seconds
0.060 Seconds
0.070 Seconds
0.060 Seconds
Select Query Results for table “HISTORY” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.044 Seconds
0.093 Seconds
0.088 Seconds
0.113 Seconds
0.173 Seconds
0.101 Seconds
0.099 Seconds
0.097 Seconds
0.130 Seconds
0.163 Seconds
0.116 Seconds
0.110 Seconds
0.145 Seconds
0.100 Seconds
0.125 Seconds
0.105 Seconds
52
Tables shown below are the results for the Join query for both databases.
Join Query Results for Oracle database:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.020 Seconds
0.020 Seconds
0.060 Seconds
0.020 Seconds
0.020 Seconds
0.040 Seconds
0.050 Seconds
0.060 Seconds
0.060 Seconds
0.060 Seconds
0.080 Seconds
0.070 Seconds
0.320 Seconds
0.100 Seconds
0.110 Seconds
0.140 Seconds
Join Query Results for SQL database:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.101 Seconds
0.178 Seconds
0.198 Seconds
0.174 Seconds
0.381 Seconds
0.178 Seconds
0.207 Seconds
0.203 Seconds
0.197 Seconds
0.244 Seconds
0.280 Seconds
0.323 Seconds
0.172 Seconds
0.175 Seconds
0.198 Seconds
0.200 Seconds
53
Tables given below are the performance measurements for Update query.
Update Query Results for table “BRANCHES” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.030 Seconds
0.001 Seconds
0.010 Seconds
0.001 Seconds
Update Query Results for table “BRANCHES” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.035 Seconds
0.016 Seconds
0.013 Seconds
0.015 Seconds
0.032 Seconds
0.053 Seconds
0.013 Seconds
0.042 Seconds
0.034 Seconds
0.019 Seconds
0.018 Seconds
0.014 Seconds
0.021 Seconds
0.013 Seconds
0.026 Seconds
0.020 Seconds
54
Update Query Results for table “TELLERS” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.020 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.030 Seconds
0.020 Seconds
0.030 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.650 Seconds
0.001 Seconds
0.020 Seconds
0.001 Seconds
Update Query Results for table “TELLERS” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.014 Seconds
0.021 Seconds
0.017 Seconds
0.024 Seconds
0.014 Seconds
0.013 Seconds
0.014 Seconds
0.045 Seconds
0.025 Seconds
0.025 Seconds
0.014 Seconds
0.014 Seconds
0.013 Seconds
0.022 Seconds
0.012 Seconds
0.014 Seconds
55
Update Query Results for table “ACCOUNTS” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.030 Seconds
0.020 Seconds
0.001 Seconds
0.010 Seconds
0.010 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.080 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
Update Query Results for table “ACCOUNTS” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.018 Seconds
0.014 Seconds
0.014 Seconds
0.015 Seconds
0.014 Seconds
0.015 Seconds
0.014 Seconds
0.043 Seconds
0.065 Seconds
0.028 Seconds
0.019 Seconds
0.033 Seconds
0.022 Seconds
0.013 Seconds
0.051 Seconds
0.045 Seconds
56
Update Query Results for table “HISTORY” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.040 Seconds
0.001 Seconds
0.030 Seconds
0.020 Seconds
0.010 Seconds
0.020 Seconds
0.070 Seconds
0.030 Seconds
0.001 Seconds
0.010 Seconds
0.050 Seconds
0.001 Seconds
0.001 Seconds
Update Query Results for table “HISTORY” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.098 Seconds
0.014 Seconds
0.025 Seconds
0.014 Seconds
0.027 Seconds
0.067 Seconds
0.039 Seconds
0.014 Seconds
0.149 Seconds
0.062 Seconds
0.059 Seconds
0.014 Seconds
0.014 Seconds
0.038 Seconds
0.065 Seconds
0.020 Seconds
57
Tables given below are performance measurements for the Delete queries on both
DBMS. Delete Query Results for table “BRANCHES” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.010 Seconds
0.001 Seconds
0.020 Seconds
0.001 Seconds
0.030 Seconds
0.001 Seconds
0.001 Seconds
0.020 Seconds
0.090 Seconds
0.001 Seconds
0.010 Seconds
0.001 Seconds
Delete Query Results for table “BRANCHES” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.020 Seconds
0.020 Seconds
0.014 Seconds
0.064 Seconds
0.015 Seconds
0.055 Seconds
0.048 Seconds
0.031 Seconds
0.042 Seconds
0.051 Seconds
0.020 Seconds
0.028 Seconds
0.014 Seconds
0.059 Seconds
0.028 Seconds
0.020 Seconds
58
Delete Query Results for table “TELLERS” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.040 Seconds
0.040 Seconds
0.020 Seconds
0.050 Seconds
0.020 Seconds
0.001 Seconds
0.001 Seconds
0.001 Seconds
0.070 Seconds
0.030 Seconds
0.040 Seconds
0.001 Seconds
2.380 Seconds
0.070 Seconds
0.001 Seconds
0.020 Seconds
Delete Query Results for table “TELLERS” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.050 Seconds
0.029 Seconds
0.034 Seconds
0.019 Seconds
0.027 Seconds
0.030 Seconds
0.051 Seconds
0.017 Seconds
0.014 Seconds
0.029 Seconds
0.028 Seconds
0.031 Seconds
0.028 Seconds
0.030 Seconds
0.026 Seconds
0.025 Seconds
59
Delete Query Results for table “ACCOUNTS” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.020 Seconds
0.020 Seconds
0.030 Seconds
0.020 Seconds
0.010 Seconds
0.010 Seconds
0.020 Seconds
0.010 Seconds
0.001 Seconds
0.001 Seconds
0.010 Seconds
0.001 Seconds
0.360 Seconds
0.020 Seconds
0.020 Seconds
0.001 Seconds
Delete Query Results for table “ACCOUNTS” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.068 Seconds
0.015 Seconds
0.016 Seconds
0.015 Seconds
0.068 Seconds
0.060 Seconds
0.014 Seconds
0.028 Seconds
0.045 Seconds
0.032 Seconds
0.484 Seconds
0.014 Seconds
0.014 Seconds
0.128 Seconds
0.053 Seconds
0.028 Seconds
60
Delete Query Results for table “HISTORY” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.001 Seconds
0.010 Seconds
0.010 Seconds
0.001 Seconds
0.001 Seconds
0.010 Seconds
0.020 Seconds
0.001 Seconds
0.020 Seconds
0.001 Seconds
0.010 Seconds
0.020 Seconds
0.220 Seconds
0.020 Seconds
0.001 Seconds
0.001 Seconds
Delete Query Results for table “HISTORY” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Query Execution Time
0.018 Seconds
0.015 Seconds
0.015 Seconds
0.015 Seconds
0.047 Seconds
0.029 Seconds
0.032 Seconds
0.020 Seconds
0.079 Seconds
0.059 Seconds
0.084 Seconds
0.014 Seconds
0.055 Seconds
0.031 Seconds
0.047 Seconds
0.040 Seconds
61
Memory usage for table “BRANCHES” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Memory Usage
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
Memory usage for table “BRANCHES” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Memory Usage
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.016 MB
0.016 MB
0.016 MB
0.016 MB
0.016 MB
0.016 MB
0.016 MB
62
Memory usage for table “TELLERS” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Memory Usage
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
Memory usage for table “TELLERS” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Memory Usage
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.016 MB
0.016 MB
0.016 MB
0.016 MB
0.016 MB
0.016 MB
0.023 MB
0.023 MB
0.023 MB
63
Memory usage for table “ACCOUNTS” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Memory Usage
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
Memory usage for table “ACCOUNTS” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Memory Usage
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.016 MB
0.016 MB
0.016 MB
0.016 MB
0.016 MB
0.016 MB
0.023 MB
0.023 MB
0.023 MB
64
Memory usage for table “HISTORY” in Oracle:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Memory Usage
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
0.06 MB
Memory usage for table “HISTORY” in SQL:
Number of Rows
20th Row
40th Row
60th Row
100th Row
150th Row
200th Row
250th Row
300th Row
350th Row
400th Row
450th Row
500th Row
550th Row
600th Row
650th Row
700th Row
Memory Usage
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.008 MB
0.016 MB
0.016 MB
0.016 MB
0.016 MB
0.023 MB
0.023 MB
0.023 MB
0.023 MB
0.031 MB
0.031 MB
0.031 MB
65
BIBLIOGRAPHY
[1] http://reference.kfupm.edu.sa/content/d/e/design_and_implementation_of_a_dbm
s_perf_2133961.pdf
[2] Technical White Paper, “MySQL Performance Benchmarks”, 2006
[3] Bitton, D., DeWitt, D.J., Turbyfill, C.: Benchmarking Database Systems A
Systematic Approach. VLDB 1983: 8-19.
[4] A.Ailamaki. “Database Architecture for New Hardware.” Tutorial. In Proc.
VLDB, 2004.
[5] P. A. Boncz, S. Manegold, and M. L. Kersten. “Database Architecture Optimized
for the New Bottleneck: Memory Access.” In Proc. VLDB, 1999.
[6] http://informingscience.org/proceedings/InSITE2006/IISITLee172.pdf
[7] http://www.theserverside.com/news/1364725/Tips-on-Performance-Testing-andOptimization
[8] http://docsdrive.com/pdfs/medwelljournals/jeasci/2008/155-160.pdf
[9] http://www.dbd.puc-rio.br/depto_informatica/04_40_macedo.pdf
[10] http://eden.dei.uc.pt/~henrique/papers/PRDC11_cr.pdf
[11] http://pages.cs.wisc.edu/~dewitt/includes/benchmarking/sigmod84.pdf
[12] http://dev.mysql.com/doc/refman/5.1/en/pluggable-storage-overview.html
[13] http://www.techotopia.com/index.php/MySQL_Database_Architecture
[14] http://www.ehow.com/about_5387852_advantages-microsoft-sql.html
66
[15] http://www.simple-talk.com/sql/database-administration/ten-common-databasedesign-mistakes/
[16] http://www.learn.geekinterview.com/database/oracle/advantages-of-usingoracle.html
[17] http://www.ehow.com/list_6309384_advantages-disadvantages-oracle-sql.html
[18] http://www.rocket99.com/techref/8690.html
[19] http://www.wikipedia.org/